Comparative analysis of background EEG activity in juvenile myoclonic epilepsy during valproic acid treatment: a standardized, low-resolution, brain electromagnetic tomography (sLORETA) study

Juvenile myoclonic epilepsy (JME) is the most common and well-defined generalized epilepsy syndrome. It usually begins in adolescence, with a peak onset between 12 and 18 years [1, 2]. JME is characterized by myoclonic seizures, especially shortly after awakening. Generalized tonic-clonic seizures often occur, and one-third of patients also have absence seizures [3]. Interictal and ictal encephalography (EEG) typically show rapid (4-6Hz) spike-waves and polyspike-waves, and photosensitivity occurs in 30% of the patients [3, 4]. Valproic acid (VPA) is the treatment drug of choice and effectively controlled all seizures in about 80 – 90% of the patients [3, 4].

By definition, the background EEG is normal in JME patients and not accompanied by other developmental and cognitive problems [1, 4]. However, some recent studies using quantitative EEG (qEEG) analysis reported abnormal changes in background activity in JME patients [5‐7]. In addition, neuropsychological studies showed cognitive dysfunction [8‐10], and several studies using multimodal methods such as magnetoencephalography (MEG), functional magnetic resonance imaging (fMRI), and evoked potentials have supported these abnormalities in JME [11‐15].

QEEG analysis is an excellent tool to evaluate background EEG and is more sophisticated, with brain-mapping techniques [5, 16, 17]. In brain mapping, the distributed source model has underlying advantages, and its algorithms address the inverse problem with few lead-in assumptions [18]. However, only a few studies have investigated the EEG characteristics of JME using a distributed model [7, 19]. Moreover, background EEG changes in patients undergoing antiepileptic drug (AED) treatment have rarely been quantified. QEEG investigation in patients undergoing AED treatment might be a useful approach to explore the electrophysiological characteristics and AED effects in JME. To better understand neurophysiological alterations and their progress in JME, the present study investigated changes in background EEG activity during treatment with VPA, the drug of choice in JME. We performed a comparative qEEG analysis of background EEG activity by a distributed source model in patients undergoing VPA treatment.

VPA decreased background EEG activity in the delta and theta frequency bands (threshold log-F-ratio = ±1.414, p < 0.05; threshold log-F-ratio= ±1.465, p < 0.01).

In the delta frequency band, comparative analysis using SnPM of the sLORETA data revealed significant current density differences in the occipital lobe (superior occipital gyrus, middle occipital gyrus, inferior occipital gyrus, cuneus, lingual gyrus, and fusiform gyrus), the parietal lobe (precuneus), and the limbic lobe (posterior cingulate) after six months of VPA treatment. In the theta frequency band, comparative analysis revealed significant current density differences in the frontal lobe (superior frontal gyrus, middle frontal gyrus, medial frontal gyrus, precentral gyrus, and sub-gyral), the occipital lobe (middle occipital gyrus, cuneus, and lingual gyrus), and the limbic lobe (anterior cingulate and posterior cingulate). Otherwise, there was no significant difference. Table 1 lists the regions with significant current density differences (p < 0.01 and p < 0.05). Figures 1 and 2 respectively show statistical color maps on cortical models and MRI images, respectively, for regions with significant differences in the delta frequency band. Figures 3 and 4 show the corresponding maps in the theta frequency band, respectively.

The maximum current density difference was found in the cuneus of the left occipital lobe in the delta frequency band (MNI coordinate [x, y, z = −15, −100, 15], Brodmann area 18) (log-F-ratio = −1.840, p < 0.01) and the medial frontal gyrus of the left frontal lobe in the theta frequency band (MNI coordinate [x, y, z = −35, 30, 45], Brodmann area 9) (log-F-ratio = −1.610, p < 0.01).

Our results showed that VPA significantly decreased the background EEG activity in low-frequency (delta-theta component) bands across the frontal, parieto-occipital, and limbic lobes.

Several studies have reported background EEG abnormalities and their progress after treatment in epilepsy [31‐33]. Some previous studies have also reported background EEG activity changes in patients with JME compared to healthy controls [6, 7, 34]. However, the anatomical extent and frequency bands involved in these studies were inconsistent. Some subsequent studies reported the effects of AED on background EEG activity in patients with JME. Santiago-Rodríguez et al. reported the analysis of background EEG activity in unmedicated and medicated patients with JME [5]. Their results showed various changes according to the anatomical location and frequency of the bands, and these changes were more evident in the frontoparietal region and unmedicated patients. Clemens et al. reported that VPA reduced delta and theta band power in background EEG activity of patients with idiopathic generalized epilepsy [25]. The anatomical distribution was located in the anterior part of the cortex, which includes the entire frontal, anterior temporal, and anterior parietal lobe, insula, and hippocampus.

Recent studies have reported the focal features of JME, predominantly in the frontal lobe [35‐37]. Neuropsychological studies also revealed subtle cognitive deficits in patients with JME, indicating frontal lobe dysfunction [10, 38]. In addition, the involvement of other regions, as well as the frontal cortex, has been reported. The sensorimotor and visual cortex areas were particularly involved. These results have been suggested in several functional studies using multimodal methods, such as EEG, MEG, fMRI, and evoked potentials [39‐41], consistent with our study. Moreover, our results revealed that AED could reduce the low-frequency band power of the background EEG activity in these regions.

Several explanations can be suggested for our results. Low-frequency background waves on EEG indicate the involvement of functional deficits in the cortical regions [25, 26]. Previous studies reported the functional impairment of JME patients due to the anatomical association of the findings. Our results are a continuation of previous studies and also suggest that antiepileptic treatment would ameliorate these dysfunctions in JME patients. Meanwhile, some studies have reported enhanced low-frequency activity in background EEGs in patients with epilepsy [23, 42, 43]. These changes were interpreted as increased excitability or synchronization that are responsible for seizures and anticonvulsants reduced this activity in the cortical areas involved. JME is a representative generalized epilepsy but generalized seizures do not always develop equally throughout the brain [44]. As noted, the focal features of JME have been reported in several studies. The areas involved were mainly the frontal lobe, and the sensorimotor and visual cortex have often been included [35, 36, 39‐41]. We found that the areas involved in those studies closely overlapped with the regions in our results, including the frontal, parieto-occipital, and limbic lobes. Furthermore, limbic lobe involvement was the distinctive feature of our results.

Our study suggested the effect of anticonvulsants on the neural networks involved in JME. To the best of our knowledge, this was the first study using qEEG analysis by the distributed model exclusively in JME. However, our results need to be elucidated in more detail by multimodal techniques, including positron emission tomography, single-photon emission computed tomography, fMRI, and neuropsychological tests. Those multimodal techniques would define the relevant areas in detail, with concepts of the epileptogenic zone, ictal onset zone, symptomatic zone, and irritative zone. In addition, case-control studies, including healthy controls, would further confirm our results.

An additional limitation of this study was the small number of patients. Larger sample size would improve the statistical power and generalizability of the results. However, the SnPM method of sLORETA, which was performed for each contrast with the built-in voxel-wise randomization tests (5000 permutations), was well-suited to our study design and sample size [16, 27]. We used EEG data with 19 electrodes for source localization. The high-density EEG with more electrodes can substantially improve results of source reconstruction. However, only a few clinics are appropriately equipped. In many situations, it is difficult to record EEG with a high number of electrodes. EEG source localization based on recordings with a limited number of electrodes can be better integrated into the clinical setting and broaden the practical applications. sLORETA localization properties have been independently validated [45] and the accuracy has been demonstrated to be similar in high- and low-density EEG [46, 47]. Therefore, we believe that there is a justification for reporting our results.

In conclusion, this study demonstrated that antiepileptic treatment reduced low-frequency (delta-theta component) background EEG activity in specific areas of patients with JME. In addition, these findings suggest the focal features and the possibility of functional deficits in JME patients. Our results, including the frontal, parieto-occipital, and limbic lobes, were consistent with previous studies and also showed distinctive features. We believe that our results contribute to further understanding the electrophysiological characteristics and AED effects of patients with JME. In the future, more expanded studies are needed including multimodal techniques, age and gender-matched controls, and a larger sample size. Such studies would confirm our results and further expand the understanding of JME.